DESCRIPTION: Applicant's Abstract This retrospective cost evaluation study is based on a six year NIDA funded research program studying effective treatments for homeless persons with co-existing Substance Use (primarily crack cocaine) and other Mental Disorders. The population was 27.7% female, 83.0% African American, mean age of 37.7 years. The goal of this research was to discover an effective treatment approach which would retain and treat homeless substance abusers. The results demonstrated positive treatment outcomes. Specific aims for this study are: 1. Conduct a retrospective cost evaluation of two related controlled studies of drug treatment for homeless persons with Substance Use and co-existing Mental Disorders conducted from 1990-93 (Homeless I) and 1994-97 (Homeless II). From Homeless I, we will compare weekly counseling (UC) -to- day treatment plus abstinent contingent housing and work therapy (EC). From Homeless II we will compare day treatment alone (DT) -to- day treatment plus abstinent contingent housing and work therapy (DT+); 2. Determine the most cost-effective treatment approach in achieving successful short and long-term verifiable outcomes on abstinence (measured by urinanalysis) and employment earnings (through secondary data linkage with the Alabama Department of Industrial Relations), and secondarily, HIV/AIDS risk; 3. Analyze important treatment utilization (e.g., length of stay) and person characteristics (e.g., sex, race, prior treatment, Mental Disorders) as predictors of achieving successful outcomes and related costs; and 4. Disseminate public policy recommendations to service delivery providers and resource decision makers. The investigators of this study are the principal investigators from Homeless I and II plus a health economist.